Trends in Mortgage-Credit Financing: Household Consumption

Lars Risbjerg, Economics

INTRODUCTION AND SUMMARY

Since the reintroduction of adjustable-rate mortgage-credit loans in 1996, a range of additional products to finance owner-occupied housing has been launched, notably deferred-amortisation loans and the banks' mortgage loans against real property as collateral. This article considers how the development in financing of owner-occupied housing can influence the households' consumption.

The households' interest payments on mortgage-credit loans have been reduced in recent years as adjustable-rate mortgage-credit loans have become more prevalent and interest rates have fallen. Viewed in isolation, this has allowed households to increase their consumption. At the same time, the households' disposable income, and presumably also their consumption, has become more exposed to future fluctuations in interest rates. On average, a rise in short-term interest rates by 1 percentage point increases the interest after tax on adjustable-rate mortgage-credit loans by almost 1 per cent of the homeowners' disposable income.

The development in the financing of owner-occupied housing can also affect private consumption since the introduction of deferred-amortisation loans and the banks' mortgage loans has weakened the link between mortgage-credit financing and subsequent loan repayments. This has made it easier for a household to achieve its preferred savings profile. The introduction of additional products can also strengthen the link between housing wealth and consumption and may lead to larger fluctuations in housing prices and thus to greater cyclical fluctuations than previously seen.

The households' borrowing for other purposes than residential investments has increased recently and is at a high level viewed over a number of years. This indicates that mortgaging of the home for other purposes then housing has been supported by the introduction of additional housing finance products. However, the households' propensity to consume does not differ from e.g. the level during the upswing in the 1990s.

INTEREST COSTS AND INTEREST EXPOSURE

Today, adjustable-rate mortgage-credit loans make up half of the households'^[1] total mortgage debt.

The households' interest payments on mortgage-credit loans have declined in recent years, even though mortgage debt is rising. Viewed in isolation, this has allowed households to increase their consumption. The growing prevalence of adjustable-rate mortgage-credit loans has contributed to reducing the interest costs of mortgage-credit borrowing since fixed long-term mortgage-credit interest rates are normally higher than the more volatile short-term interest rates. Moreover, the rate of interest on adjustable-rate mortgage-credit loans has fallen with the level of interest rates, which has also made it possible to remortgage fixed-rate loans to a lower interest rate.

Interest payments by the households to banks and mortgage-credit institutes totalled kr. 47 billion after tax in 2005, of which interest payments to mortgage-credit institutes accounted for kr. 36 billion, equivalent to 4.6 per cent of disposable income.^[2] For comparison, interest costs on mortgage-credit loans in, respectively, 2003 and 2004 were 5.4 and 5.0 per cent of disposable income. On the other hand, the households' interest income has declined. However, the households only to a limited extent directly own mortgage-credit bonds, while indirect ownership, especially via pension savings, is somewhat higher. The net effect has been falling net interest costs, particularly if pension savings, which cannot be disposed of immediately, are omitted.

The more widespread use of adjustable-rate loans has at the same time made the households' disposable income, and thus presumably also private consumption, more sensitive to future fluctuations in short-term interest rates.^[3] The households are also exposed to changes in the short-term interest rates via their bank loans, cf. Chart 1.

On a permanent increase in short-term interest rates by 1 percentage point, the households' annual interest costs after tax on their existing adjustable-rate mortgage-credit loans will increase by kr. 4 billion. This is equivalent to 0.6 per cent of the households' disposable income. Homeowners' income constitutes an estimated two thirds of the households' total income after tax. So the increase in interest costs is equivalent to almost 1 per cent of the homeowners' disposable income.^[4]

However, an increase in interest rates will also entail higher interest income from the households' investments in variable-rate assets, cf. above, and it takes some time for an increase in interest rates to pass through to all adjustable-rate mortgage-credit loans. Around 29 per cent of the total outstanding volume of adjustable-rate mortgage-credit loans is subject to adjustment after more than 1 year. In addition, the interest rates on some of the adjustable-rate mortgage-credit loans are capped, which protects borrowers against large interest-rate increases. Capped mortgage-credit bonds constitute around 26 per cent of total outstanding adjustable-rate mortgage-credit loans. For adjustable-rate mortgage-credit annuity loans subject to ongoing repayment, the near-term repayments are reduced if interest rates rise. For deferred-amortisation adjustable-rate loans, an increase in interest rates has a full impact on debt service payments.

MORTGAGE-CREDIT BORROWING AND MORTGAGING OF HOUSING WEALTH

Private consumption tends to reflect the development in the households' wealth, which notably comprises housing wealth. By accumulating and spending wealth, the households can achieve an intertemporal reallocation of consumption compared to income. For instance, if the households' wealth increases, they can realise some of the increase as consumption today. In addition to the wealth effect, housing wealth can be used as collateral in order to borrow for consumption purposes. Rising housing prices can underpin consumption by increasing mortgageable values and alleviating credit restrictions.

The households' mortgage-credit loans and total borrowing from banks and mortgage-credit institutes as a ratio of their disposable income has more than doubled in the last 25 years, cf. Chart 2. The increase is primarily related to falling interest rates and rising housing prices, but also to the liberalisation of mortgage credit.

The development in home financing has made it gradually easier to mortgage the home for other purposes than housing. For example, loans with deferred amortisation, which were introduced in 2003, have led to greater separation of decisions concerning, respectively, mortgage-credit financing and savings. To the extent that the households have had limited access to liquidity, this may have contributed to increased borrowing and consumption.

Deferred-amortisation loans constitute around 30 per cent of lending by the mortgage-credit institutes to the households. Homeowners' total ordinary repayments as a ratio of their outstanding mortgage debt has thus declined after the introduction of deferred-amortisation loans. In nominal terms, total ordinary repayments by homeowners are, however, by and large unchanged because the households have increased their mortgage debt. Ordinary repayments on loans for owner-occupied housing and holiday cottages totalled kr. 19 billion in 2005. For the individual homeowner, the opportunity to separate borrowing and savings decisions entails a risk of short-term focus on the first-year payments on the mortgage-credit loan, without adequately taking the consequences of lower savings for future consumption possibilities into account.

The banks' mortgage loans function as overdraft facilities against the home as collateral and are thus a mortgage-credit-type product. These loans also make it possible to postpone repayments on the debt.

In a European context, Denmark has a well-developed mortgage-credit market in terms of remortgaging opportunities and supplementary mortgage credit, as well as its range of mortgage-credit products, e.g. various types of interest-rate adjustment and repayment profiles.^[5] Empirical studies indicate that a well-developed mortgage-credit market with good opportunities for borrowing against the free mortgageable value of owner-occupied housing increases mortgage debt and strengthens the relation between consumption and housing wealth.^[6] The relation between housing prices and the consumption ratio is illustrated in Chart 3.

Just as new home financing products can influence consumption demand, they can also influence housing demand and housing prices, which in turn may affect consumption.^[7] Moreover, there are indications that housing prices – and thus housing wealth – are more volatile in countries where variable-rate housing loans are most prevalent than in countries with predominantly fixed-rate housing loans. Likewise, the impact of short-term interest rates on housing prices tends to be greater.^[8] Consequently, the development of new products may entail larger

fluctuations in housing prices, and thereby greater cyclical fluctuations, than previously, if housing wealth is significant to private consumption and residential investments.^[9]

Borrowing by the households and housing prices are typically procyclical, cf. Chart 4. For instance, if housing prices flatten or decrease during a cyclical slowdown, the households often reduce their borrowing against the home as collateral. This can reduce consumption demand and amplify the cyclical fluctuations. Mortgaging of housing wealth for other purposes than housing (mortgage equity withdrawal, cf. below) and the risk that the contribution to economic growth from borrowing is eliminated or changes sign has attracted much attention in a number of countries in recent years.

This applies particularly in the USA, where borrowing against the home as collateral has increased in connection with rising housing prices.^[10]

Mortgage equity withdrawal and consumption
One way of finding out how mortgage-credit borrowing affects private consumption is to ask homeowners. In connection with remortgaging, homeowners were interviewed about their increased mortgage-credit borrowing and how the proceeds are used, cf. Box 1. According to these surveys, in 2004 homeowners primarily used the additional mortgage credit for residential investments, while a small proportion was used to increase consumption. These results should, however, be interpreted with caution since the households' demand in the absence of remortgaging is not known. In addition, it is difficult to draw any conclusions about trends as surveys are only available from two years.

An alternative approach to assessing the significance of mortgage-credit borrowing to consumption is to consider mortgage equity withdrawal, defined as the households' mortgage-credit borrowing less residential investments.^[11] Mortgage equity withdrawal is an expression of the homeowners' dissaving in their home that is used for consumption, but also to reduce other debt and acquire financial assets.^[12] Homeowners' supplementary borrowing, e.g. in connection with remortgaging, is a source of the households' borrowing and mortgage equity withdrawal, which also arise in connection with purchase and sale of homes, cf. Box 2.

THE HOUSEHOLDS' MORTGAGE-CREDIT BORROWING	Box 2
The households can often raise supplementary mortgage credit in connection with remortgaging. The turnover of housing may also increase the overall mortgage-credit borrowing by the households, since the seller's outstanding mortgage debt is typically lower than the buyer's new mortgage-credit loan. This effect is underpinned by price increases in the housing market. Consequently, the households' mortgage-credit borrowing reflects both the turnover in the housing market and early redemptions. The latter arise both in connection with remortgaging, and when properties are bought and sold, cf. the Chart below. Presumably any effects of new products on borrowing and private consumption can primarily be identified in connection with remortgaging and other supplementary borrowing. In a statistical context, the sources of borrowing and mortgage equity withdrawal are not directly identifiable.
MORTGAGE-CREDIT BORROWING, EARLY REDEMPTIONS AND TURNOVER OF HOUSING
Note: Early redemptions are 4-quarter averages, and the first observation is from 1994. The number of trades relates to trades within one year, and the most recent observation is from 2004. Source: Statistics Denmark, Ministry of Finance, Copenhagen Stock Exchange and Danmarks Nationalbank.

Substitution between bank and mortgage-credit loans takes place on an ongoing basis. In addition to mortgage credit, the households raise bank loans against the home as collateral. Most of the growth in the banks' lending in recent years relates to lending for housing purposes, including banks' mortgage loans.^[13] Mortgage equity withdrawal, both including and excluding bank loans, increased in 2005, cf. Chart 5. It is difficult to state whether the recent increase is a result of product development in the mortgage-credit sector, or of housing prices and cyclical factors. If mortgage equity withdrawal has risen to a higher level than previously, this could indicate that the introduction of additional products has had an effect. Mortgage equity withdrawal is now somewhat above the level towards the end of the upswing in the 1990s, particularly if bank lending is included. The households' total borrowing from banks and mortgage-credit institutes, which can be used for consumption and net financial savings, has reached the same level as during the boom in the mid-1980s.

A proportion of the most recent increase in mortgage equity withdrawal presumably reflects supplementary borrowing in connection with the increasing remortgaging activity. The introduction of new house financing products may have stimulated remortgaging activity. In the early 1990s, mortgage-credit borrowing and mortgage equity with drawal were also underpinned by the development in mortgage-credit

financing. In 1992 the opportunities to raise supplementary mortgage credit were enhanced. From 1993, it was no longer compulsory to use "mixed loans" for home financing, and 30-year annuity loans were reintroduced, leading to a wave of remortgaging when interest rates subsequently fell.^[14]

To what extent has mortgage equity withdrawal been used for consumption? In connection with rising housing prices and housing wealth the consumption ratio has increased during the current upswing. However, the consumption ratio does not differ from the level towards the end of the upswing in the 1990s. As was the case during the upturn in the 1990s, the households' financial savings surplus (net borrowing/net lending) has been reduced during the upswing that began in 2003, cf. Chart 6. However, the households' financial savings deficit is smaller than their borrowing from banks and mortgage-credit institutes since the households taken as one have invested part of the mortgage equity withdrawal in financial assets, e.g. via pension contributions. The increase in the households' pension wealth (including capital gains) since the end of 1998 by and large corresponds to the growth in their loans from banks and mortgage-credit institutes. At the same time, the households' net wealth has been increasing in recent years. The accumulation of wealth has been driven primarily by capital gains on housing and shares. Viewed in isolation, this has improved the households' future consumption possibilities. Analyses of financial stability also indicate that in general the households are well-consolidated.^[15]